Calculate ADC Resolution
Enter converter limits and front-end scaling. Leave ENOB empty to calculate with nominal resolution. Enter signal RMS only when you need signal-to-quantization-noise ratio.
Example Data
These examples use ideal converter behavior. They show how bit depth and voltage span change the input-referred step.
| Bits | Range | Reference | Gain | Input-Referred LSB | Resolution |
|---|---|---|---|---|---|
| 10 | Unipolar | 3.3 V | 1 | 3.2227 mV | -49.84 dBV |
| 12 | Unipolar | 3.3 V | 1 | 0.8057 mV | -61.87 dBV |
| 12 | Unipolar | 3.3 V | 4 | 0.2014 mV | -73.91 dBV |
| 16 | Bipolar | 2.5 V | 1 | 76.2939 µV | -82.35 dBV |
Formula Used
The calculator starts with the ADC span. It then applies the selected bits, gain, attenuation, and optional ENOB.
When ENOB is entered, the calculator uses ENOB for practical resolution and noise estimates. Headroom only reduces the recommended usable full-scale sine value. It does not change the physical LSB.
How to Use This Calculator
- Enter the nominal converter bit depth.
- Choose unipolar or bipolar operation.
- Enter the reference voltage used by the converter.
- Add analog gain and any attenuation before conversion.
- Enter practical ENOB when measured converter performance is known.
- Reserve headroom for peaks, offset, and tolerances.
- Enter a signal RMS value to calculate its dBV and SQNR.
- Read the input-referred LSB first. It represents one digital code at the original source.
ADC Resolution Fundamentals
Understanding ADC Resolution
ADC resolution describes the smallest voltage change a converter can separate. Bit depth sets the number of available digital codes. A larger code count creates a smaller voltage step. That step is commonly called one least significant bit, or LSB. This calculator expresses the step in volts and dBV. dBV uses one volt RMS as its reference. A value below one volt produces a negative dBV value. This makes very small steps easier to compare. The result is useful when selecting a converter or scaling an analog front end.
Resolution alone does not guarantee accuracy. Reference stability matters. Input noise matters. Layout and grounding also matter. Treat calculated figures as ideal design targets. Confirm final performance with measurements across temperature, frequency, operating conditions, intended signal ranges, and supply variations.
From Bit Depth to Voltage Steps
An ideal converter with N bits has 2 raised to N possible codes. Divide the full input span by that code count. The answer is the ideal LSB size at the converter input. A unipolar converter usually spans zero to its reference voltage. A bipolar converter spans negative reference voltage to positive reference voltage. Therefore, its total span is twice the reference voltage. The calculator applies the selected range automatically. It then converts the LSB voltage to dBV with twenty times the base ten logarithm.
Input-Referred Resolution
Converter-side resolution is not always the same as source-side resolution. A gain stage magnifies the source before conversion. This lets one code represent a smaller source voltage. An attenuator does the opposite. It enlarges the input-referred voltage represented by each code. The calculator divides by gain and multiplies by attenuation. It reports the result as an input-referred LSB. Use this value when judging whether a sensor change can be detected. Check that the gain stage does not cause clipping near the converter limits.
Quantization Noise and ENOB
Quantization creates a small uncertainty around each code boundary. For an ideal uniform converter, its RMS value is one LSB divided by square root of twelve. The calculator reports this noise in volts and dBV. It also estimates signal-to-quantization-noise ratio when an input RMS signal is supplied. Real devices include thermal noise, distortion, clock noise, and reference errors. Effective number of bits, or ENOB, captures many of those losses. Enter ENOB when measured performance is lower than nominal bit depth. The calculator then uses it for practical resolution estimates.
Applying the Result
Use the recommended full-scale sine value as a setup check. It includes the headroom selected for overload protection. Headroom reduces the largest permitted signal, not the physical code size. Compare expected sensor changes with the input-referred LSB and noise level. Prefer a signal change that is comfortably larger than both. Use dBV when comparing stages with different voltage levels. Use dBFS when comparing a signal against converter full scale. Keep units consistent. Review offset, common-mode limits, bandwidth, and sampling effects before finalizing a design.
Frequently Asked Questions
What does dBV mean?
dBV expresses a voltage relative to 1 V RMS. Zero dBV equals 1 V RMS. Values below 1 V RMS are negative. It is useful for comparing small voltage steps across analog stages.
Why does the calculator use 2 raised to N?
An ideal N-bit converter has 2^N possible code levels. Dividing the total input span by those levels gives the ideal step size. Some data sheets use endpoint conventions, but this calculator uses the standard ideal LSB relationship.
Should I use nominal bits or ENOB?
Use nominal bits for an ideal estimate. Use ENOB when you have measured data-sheet or laboratory performance. ENOB gives a more practical estimate because it reflects noise and distortion losses.
Does headroom change ADC resolution?
No. Headroom reduces the signal amplitude you choose to use. The physical code width remains unchanged. It helps avoid clipping from peaks, offsets, reference drift, and component tolerance.
How does amplifier gain affect input resolution?
Gain reduces the input-referred LSB because a smaller source voltage reaches the same ADC code change. Excessive gain can clip the signal. Check the full input range before increasing gain.
What attenuation factor should I enter?
Enter 1 when there is no attenuator. Enter 10 for a 10:1 divider. The calculator multiplies the ADC-side LSB by this factor to refer it back to the original source.
Why is the quantization noise value lower than one LSB?
Ideal quantization error is modeled as uniformly distributed across one LSB. Its RMS value is LSB divided by square root of twelve. That RMS value is lower than the full code width.
Can this calculator measure actual ADC accuracy?
No. It estimates resolution and ideal quantization limits. Actual accuracy also depends on offset, gain error, INL, DNL, reference quality, noise, temperature, and signal bandwidth.
Is dBV the same as dBFS?
No. dBV references 1 V RMS. dBFS references the converter full-scale level. Use dBV for physical voltage comparisons. Use dBFS for digital full-scale comparisons.
Why is a bipolar span twice the reference voltage?
A bipolar converter spans from negative reference voltage to positive reference voltage. The total distance between those limits is two times the reference magnitude. That wider span changes the LSB calculation.
What signal value should I enter?
Enter the RMS voltage at the original external input. The calculator uses it to report signal dBV and signal-to-quantization-noise ratio. Leave it blank when only resolution is needed.